I must have been ignoring you
I think it was when you said "positive displacement" about the vacuum pump.
Non-destructive Ruffles® Bypass Valve testing
- Thread starter Hatsuwr
- Start date
I think it was when you said "positive displacement" about the vacuum pump.
I commend you for attempting this, but unfortunately it’s not going to get to the “Elephant in the Room”….
How does this particular bypass leak impact filtration…in a working filter?…not on paper. In my humble opinion, short of an ISO, the answer could be had by having a BR particle Endurance filter test performed AGAIN, only with a flashlight test post procedure. If it’s a leaker as I suspect, then there should be a follow up test with the same filter type which would be “sealed” in the potential leak area with epoxy prior (no cutting), and test again for filtration with a post flashlight test to confirm non leakage. THEN and only then, can we compare filtration of leaker vs non leaker in the same model.
True, we would still be in the dark as correlations between magnitude of leak vs
any corresponding declines in filtration, but we could still get a reasonable assessment how this specific bypass leak impacts filtration in any material way.
Last edited:
Bypass leaking has at least two effects. First, it lowers the efficiency of the filter at all particle sizes by an amount that is proportional to the fraction of flow that goes through the leaks. Importantly, that means there is no longer any ~100% filtration rate for a reasonable particle size. If you check out the Ascent testing, the Fram, RP, AC-Delco, and Purolator filters were all nearly 100% effective for particles above 35 microns. That would not be true for a leaky filter (and actually, looking at the graph again... I wonder if the Wix was leaking.)
Second, it means that startup contaminant surges will be much worse. For a given filter diameter, these effects will be worse for shorter filters.
The magnitude of these effects is an interesting question, but not one that is going to be answered by this test. This test also won't tell you if a filter is leaking or not.
What this test can do is help identify the filter that likely has the least leakage from a batch, and I think we can all at least agree that less leaking is better.
@Ronn I posted this is another thread and nobody commented. I thought you would be all over it! Note I have inserted a second yellow line to reflect where the Fram Endurance would be with 10% volume of oil bypassing the filter element.
Mathematically speaking, it is true (I did a spreadsheet to be sure), that if 10% of the oil goes through a permanent bypass, then for a filter that was near 100% efficient, you can say that it is now 90% efficient.
The issue is that it drops 10% across the entire micron range.
Still it's better than a WIX XP, so there is some truth to the new saying in this forum "If you're going to pick a leaky filter, choose a high efficiency leaky filter".
True,
But this still begs the question, is this particular bypass leak a “deal breaker”….enough for the Endurance Filter (Ultra..Amsoil..RP…etc) to relinquish its status ask a top performing filter?
“Is a A Samurai Sword which has been dulled, still superior to all those fully sharpened replacements?”
Some vacuum pumps are positive displacement ... do some research.I must have been ignoring you
I think it was when you said "positive displacement" about the vacuum pump.
It's enough for me. Unless my understanding of things changes before my next oil change (or unless I can find some of those OG Titaniums), I'll be going with a different filter.
Also, I just realized I had a fourth XG3600. This one is from RockAuto, while the others were from Walmart. It has the same date code as filters 1 and 2: A40462.
I ran a test with filter 1, three tests with filter 4, then another test with filter 1 to make sure I was still getting results consistent with the first run. Filter 4 is in yellow. It fell right in between filters 1 and 2.
This is interesting, because it gives more evidence of performance correlating with manufacture date, even with identical appearance. The newer filter 3 performed best.
Then why even conduct it? All you're doing is comparing the vacuum level of the same filter brand/model to each other without knowing anything about what's going on inside them. If you tested a modified filter to make it a 100% non-leaker (a baseline control filter) then you might have something that would be more informative.
Like said before ... the one that's the "best" could still be leaking like a sieve compared to a 100% non-leaker.
Probably because you offered no explanation (like you did this time) of what you did and what it represented.
All you had to do was make it to the end of the sentence lol. Surely you can at least agree that less leaking is better than more leaking?
Here's a little inside info for this discussion. An ISO test lab bought some WIX XP/NAPA Platinum filters back when the manufacturer claimed the efficiency was 50% @ 20u. After the bad press got out, they removed the efficiency off their website, and if emailed or called they told you (I did this myself) that they efficiently was "proprietary".
The lab tested them per ISO 4548-12 just as they were bough off the shelf, and indeed they came in around that efficiency. They then cut some new ones open and sealed up everything except for just the media cartridge, and re-tesed them per ISO 4548-12 again, and the efficiency was around 90% @ 20u after fixing the internal lead(s). So internal leaks can certinaly impact the overall efficiency as they demonstarted.Well, all of them being within a needle width of each other on the vacuum gauge may not mean much if anything really ... some of that could be caused by other factors, like slight differences in total media area, or the bypass valves themselves are seated slightly different off the assembly line. So it's hard to distinguish the very slight delta vacuum level cause without knowing that other factors mighty be involved. If you knew the vacuum level on a non-leaker, it might give more insight to just how much these are leaking. What's the delta vacuum between a leaker and a non-leaker. You will never know unless you test for it.
Last edited:
You would have the same variability in media regardless of leaks. The assumption of media consistency is unfortunately necessary.
If one of these filters happened to have a tear in it and came in at half the pressure differential compared to the others, you aren't going to come back with "Useless test, needs a baseline," right lol? So even if you disregard the small differences revealed in the filters I have on hand (which is understandable since, as I said, this was a quick setup that would benefit from a higher flow rate), there is still value in identifying major defects like media tears before installation.
Do you have another way to identify a tear in a filter you want to end up using?
It was a test! Btw, it's still going to filter down to 5 microns better than most other filters .....
I knew about the third party testing but didn't remember how much it improved once they fixed the filter. 90% at 20 microns is quite the improvement.
If I was going to make a baseline control non-leaker filter, I would first test it before modifying to get that reading. Then I'd modify it to be 100% sealed around the leaf spring to end cap interface. Then retest it to see the effect of sealing the leaf spring. Everything on that filter would be exactly the same except for the leaf spring sealing area. If you see a decent change in the vacuum level, then you at least have a delta that correlated to only the leaf spring seal.
That baseline filter could also be used to compare other make/model of the same against it to detect leakage levels on those filters.
Having a tear in the media on an new unused filter is pretty unheard of. Something I would never worry about. I dropped the media tearing Purolators like a hot potato when that all hit the fan, lol.
A leaky leaf spring like seen here recently ... yeah, I don't think I'd use them since my calculations on @Glenda W. example comes in at 16% leakage.
Well of course ... like the meme says:
I just looked at the correspondence info I had, and it went from approx 50% @ 20u to approx 90 @ 20u.
Quick note on the FE tests. I just noticed that FE3600 #2 has longer slots in the central core than 1 and 3. The slots on 2 are more like those found on the XG and FS filters I have. This could be contributing to the difference in pressure differentials.
This next batch is interesting. These are FS3600s from AAP. Filter 1 has holes in the center core with no wire backing. Filters 2 and 3 have the spiral slots and do seem to have wire backing. This does mean that a comparison between 1 and 2/3 need to be taking with a grain of salt due to the different construction.
Dates:
Filter 1: A30202 - 2023, 020nd day
Filter 2: A40673 - 2024, 067th day
Filter 3: A40673 - 2024, 067th day
The FEs sit between about 10.5-11.5 kPa, while the XGs are 12-13 kPa. FS1/2 are right around 9.5 kPa, and FS3 is about 10.5 kPa.
I'm surprised to see 1/2 so close to each other considering their different construction, while 1/3 performed quite differently despite being made the same day. I would assume from this that filter 1 leaks significantly more than filter 3. Hard to judge filter 2 here. With no wire backing, is 2 still considered one of the OG Titaniums?
This next batch is interesting. These are FS3600s from AAP. Filter 1 has holes in the center core with no wire backing. Filters 2 and 3 have the spiral slots and do seem to have wire backing. This does mean that a comparison between 1 and 2/3 need to be taking with a grain of salt due to the different construction.
Dates:
Filter 1: A30202 - 2023, 020nd day
Filter 2: A40673 - 2024, 067th day
Filter 3: A40673 - 2024, 067th day
The FEs sit between about 10.5-11.5 kPa, while the XGs are 12-13 kPa. FS1/2 are right around 9.5 kPa, and FS3 is about 10.5 kPa.
I'm surprised to see 1/2 so close to each other considering their different construction, while 1/3 performed quite differently despite being made the same day. I would assume from this that filter 1 leaks significantly more than filter 3. Hard to judge filter 2 here. With no wire backing, is 2 still considered one of the OG Titaniums?
No
On a sidenote ... the guys at Fram that designed the OG Ultra (and it also the same guts in the Titanium) knew what to do in order to achieve high efficiency and high capacity. They used that fiber seal on the end cap to seal the leaf spring for a distinct reason ... to make that seal right in order to achieve higher efficiency.
Last edited:
So what's the progression?
Holes + wire backing -> Holes w/o wire backing -> Louvers?
